Quickstart
Install the CLI, get an answer, and know what to do when a step does not work. Budget ten minutes.
If you only want to see it work, you do not need this page yet. Go to saylek.com, enter your email, and open the sign-in link that comes back. Home shows the models on your own machines and on the machines people have shared with you. A member who has neither yet is offered two ways in: connect a machine of your own, or ask someone you trust to share theirs. Come back here to install the CLI or set up a machine of your own.
Saylek lets you use the AI models on your own machines and on machines shared with you by people you trust. Share a machine of yours and the people you choose can use its models too. Sharing is a separate decision in each direction, and installing shares nothing on its own. Personal use is free. A request that goes through your machine and completes leaves a signed record of where it ran.
Before sending sensitive content, check where it may run. A request routed through Saylek passes through its servers and is handled by the machine that serves it: your own machine, a machine shared with you, or, if you joined a group, a machine in that group. Privacy and egress explains the boundaries and local-only controls.
Before you begin
| You need | Detail |
|---|---|
| A supported machine | Released binaries target Linux x86_64 and macOS arm64 (Apple Silicon). On Windows, the Linux build runs inside WSL2 (x86_64). |
| An account | Create one at saylek.com with your email. It opens straight away. No invitation is needed to create an account: an invitation is how someone offers you their machine, it is not the door to Saylek. |
| Your install command | Below, in full. You can also open Connect a machine in Machines for guided setup. |
| Outbound HTTPS | The installer fetches a signed release; the daemon reaches Saylek over HTTPS. |
| A GPU, or a model runtime already running | Only if you want to share a machine. Using a machine somebody shared with you works fine with no GPU at all. |
You do not need Docker, a Rust or Python toolchain, or sudo. Everything runs from the
released binary.
1. Install
curl -fsSL https://saylek.com/install.sh | sh
Copy it as written. There is no token in it and nothing to fill in: the script is public, and it is the same one for everybody.
This verifies the release signature, installs saylek under ~/.saylek/, and registers a
user-level service that keeps Saylek running. It does not set up a model or share anything:
step 3 covers models.
To sign in and link this machine to your account at the same time, so it appears in
Machines, add the --connect flag. It also connects a model runtime it finds
running on this machine. It is the same script either way, and sharing stays off until you
turn it on:
curl -fsSL https://saylek.com/install.sh | sh -s -- --connect
Connect a machine in Machines walks you through the same step. To let people you choose use the machine later, see Sharing your models.
Prefer to read before you run? The URL serves a plain shell script, so fetch it, read it, then run it:
curl -fsSL https://saylek.com/install.sh -o saylek-install.sh
less saylek-install.sh
sh saylek-install.sh
You know it worked when saylek --version prints a version.
2. Check the daemon is up
saylek status
You know it worked when status reports the daemon running. Add --watch to live-tail
the state while something is settling.
If it is not running, start it with saylek start. Note that saylek start runs in the
foreground and holds the terminal until you stop it, which surprises people who expect it
to return to the prompt. Either open a second terminal for the commands below, or background
it with saylek start &. On a normal install a service manager keeps the daemon up for you,
so you should rarely need to run it by hand at all.
3. Ask your first question
saylek call "explain what a GPU does, in one sentence"
You know it worked when you get an answer back.
If nothing is set up to answer yet, saylek call says so in plain language and points at
the next step, such as saylek start when the daemon is down or saylek gpu wizard when
no model is loaded. It does not dead-end. Run it with no prompt for a built-in one-line
smoke test.
Already running Ollama, LM Studio, llama.cpp or vLLM on this machine? Connect it, and its models answer here:
saylek model add http://127.0.0.1:11434 # your runtime's address; this is Ollama's
saylek gpu wizard also offers to connect a runtime it finds running.
4. Pick where you go next
You are running. Choose by what you came for:
- Use models from an app: Connect your app for the hosted Saylek API, with OpenAI-compatible and Anthropic Messages setup guides.
- Share your machine: Sharing your models.
- Bring someone in: Invite flow.
- Understand how it works: Concepts.
If something went wrong
Common first-run stops, with the full list on Troubleshooting:
| Symptom | Likely cause | Fix |
|---|---|---|
model_not_found on your first call | Daemon is up but no model is loaded yet | saylek gpu wizard, or saylek model add <url> for a runtime already running on this machine |
saylek: command not found after install | ~/.saylek/bin is not on your PATH | Open a new shell, or add it to your shell profile |
| Calls fail with nothing available to serve them | No machine you can use serves the requested model right now, or every one that does is busy | Check saylek status and the models available in Home; retry when one is available |
| Everything fails after enabling local-only | Local-only means local, including when local cannot answer. It stops Saylek routing to anyone else; it does not override a proxy upstream you configured yourself | Unset local_only, or load a model you can serve |
Uninstalling
Two different scopes, and it matters which you mean:
saylek wipe # remove this device's local files only
saylek settings account delete # leave Saylek entirely (account: 90-day soft delete; this device's keys: removed)
saylek wipe says nothing about the server: your account and who you share with remain. To step away
without deleting anything, saylek settings account deactivate is fully reversible. The
CLI reference has the whole table.
Next steps
- Concepts: how sharing works between two people, and receipts.
- Privacy and egress: what leaves your machine.
- Troubleshooting: when a step does not behave.
Last checked 2026-09-24 · read as markdown at /docs/quickstart.md